I have one huge file (over 6GB) and about 1000 patterns. I want extract lines matching each of the pattern to separate file. For example my patterns are:
1
2
my file:
a|1
b|2
c|3
d|123
As a output I would like to have 2 files:
1:
a|1
d|123
2:
b|2
d|123
I can do it by greping file multiple times, but it is inefficient for 1000 patterns and huge file. I also tried something like this:
grep -f pattern_file huge_file
but it will make only 1 output file. I can't sort my huge file - it takes to much time. Maybe AWK will make it?
awk -F\| 'NR == FNR {
patt[$0]; next
}
{
for (p in patt)
if ($2 ~ p) print > p
}' patterns huge_file
With some awk implementations you may hit the max number of open files limit.
Let me know if that's the case so I can post an alternative solution.
P.S.: This version will keep only one file open at a time:
awk -F\| 'NR == FNR {
patt[$0]; next
}
{
for (p in patt) {
if ($2 ~ p) print >> p
close(p)
}
}' patterns huge_file
You can accomplish this (if I understand the problem) using bash "process substitution", e.g., consider the following sample data:
$ cal -h
September 2013
Su Mo Tu We Th Fr Sa
1 2 3 4 5 6 7
8 9 10 11 12 13 14
15 16 17 18 19 20 21
22 23 24 25 26 27 28
29 30
Then selective lines can be grepd to different output files in a single command as:
$ cal -h \
| tee >( egrep '1' > f1.txt ) \
| tee >( egrep '2' > f2.txt ) \
| tee >( egrep 'Sept' > f3.txt )
In this case, each grep is processing the entire data stream (which may or may not be what you want: this may not save a lot of time vs. just running concurrent grep processes):
$ more f?.txt
::::::::::::::
f1.txt
::::::::::::::
September 2013
1 2 3 4 5 6 7
8 9 10 11 12 13 14
15 16 17 18 19 20 21
::::::::::::::
f2.txt
::::::::::::::
September 2013
1 2 3 4 5 6 7
8 9 10 11 12 13 14
15 16 17 18 19 20 21
22 23 24 25 26 27 28
29 30
::::::::::::::
f3.txt
::::::::::::::
September 2013
This might work for you (although sed might not be the quickest tool!):
sed 's,.*,/&/w &_file,' pattern_file > sed_file
Then run this file against the source:
sed -nf sed_file huge_file
I did a cursory test and the GNU sed version 4.1.5 I was using, easily opened 1000 files OK, however your unix system may well have smaller limits.
Grep cannot output matches of different patterns to different files. Tee is able to redirect it's input into multiple destinations, but i don't think this is what you want.
Either use multiple grep commands or write a program to do it in Python or whatever else language you fancy.
I had this need, so I added the capability to my own copy of grep.c that I happened to have lying around. But it just occurred to me: if the primary goal is to avoid multiple passes over a huge input, you could run egrep once on the huge input to search for any of your patterns (which, I know, is not what you want), and redirect its output to an intermediate file, then make multiple passes over that intermediate file, once per individual pattern, redirecting to a different final output file each time.
Related
I have a file which has n number of rows, i want it's data to be distributed in 7 files as per below order
** my input file has n number of rows, this is just an example.
Input file
1
2
3
4
5
6
7
8
9
10
11
12
13
14
1
5
16
17
.
.
28
Output file
1 2 3 4 5 6 7
14 13 12 11 10 9 8
15 16 17 18 19 20 21
28 27 26 25 24 23 22
so if i open the first file it should have rows
1
14
15
28
similarly if i open the second file it should have rows
2
13
16
27
similarly output for the other files as well.
Can anybody please help, with below code it is doing what is required but not in required order.
awk '{print > ("te1234"++c".txt");c=(NR%n)?c:0}' n=7 test6.txt
1 2 3 4 5 6 7
8 9 10 11 12 13 14
15 16 17 18 19 20 21
22 23 24 25 26 27 28
EDIT: Since OP has changed sample of Input_file totally different so adding this solution now, again this is written and tested with shown samples only.
With xargs + single awk: (recommended one)
xargs -n7 < Input_file |
awk '
FNR%2!=0{
for(i=1;i<=NF;i++){
print $i >> (i".txt")
close(i".txt")
}
next
}
FNR%2==0{
for(i=NF;i>0;i--){
count++
print $i >> (count".txt")
close(i".txt")
}
count=""
}'
Initial solution:
xargs -n7 < Input_file |
awk '
FNR%2==0{
for(i=NF;i>0;i--){
val=(val?val OFS:"")$i
}
$i=val
val=""
}
1' |
awk '
{
for(i=1;i<=NF;i++){
print $i >> (i".txt")
close(i".txt")
}
}'
Above could be done with single awk too will add xargs + awk(single) solution in few mins too.
Could you please try following, written and tested with shown samples in GNU awk.
awk '{for(i=1;i<=NF;i++){print $i >> (i".txt");close(i".txt")}}' Input_file
The output file counter could descend for each second group of seven:
awk 'FNR%n==1 {asc=!asc}
{
out="te1234" (asc ? ++c : c--) ".txt";
print >> out;
close(out)
}' n=7 test6.txt
$ ls
file tst.awk
$ cat tst.awk
{ rec = (cnt % 2 ? $1 sep rec : rec sep $1); sep=FS }
!(NR%n) {
++cnt
nf = split(rec,flds)
for (i=1; i<=nf; i++) {
out = "te1234" i ".txt"
print flds[i] >> out
close(out)
}
rec=sep=""
}
.
$ awk -v n=7 -f tst.awk file
.
$ ls
file te12342.txt te12344.txt te12346.txt tst.awk
te12341.txt te12343.txt te12345.txt te12347.txt
$ cat te12341.txt
1
14
15
28
$ cat te12342.txt
2
13
16
27
If you can have input that's not an exact multiple of n then move the code that's currently in the !(NR%n) block into a function and call that function there and in an END section.
This might work for you (GNU sed & parallel):
parallel 'echo {1}~14w file{1}; echo {2}~14w file{1}' ::: {1..7} :::+ {14..8} |
sed -n -f - file &&
paste file{1..7}
Create a sed script to write files named filen where n is 1 thru 7 (see above first set of parameters in the parallel command and also in the paste command).
The sed script uses the n~m address where n is the starting address and m is the modulo thereafter.
The distributed files are created first and the paste command then joins them all together to produce a single output file (tab separated by default, use paste -d option to get desired delimiter).
Alternative using Bash & sed:
for ((n=1,m=14;n<=7;n++,m--));do echo "$n~14w file$n";echo "$m~14w file$n";done |
sed -nf - file &&
paste file{1..7}
I have the following test file:
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
I want to separate it in a way that each file contains the last line of the previous file as the first line. The example would be:
file 1:
1
2
3
4
5
file2:
5
6
7
8
9
file3:
9
10
11
12
13
file4:
13
14
15
16
17
file5:
17
18
19
20
That would make 4 files with 5 lines and 1 file with 4 lines.
As a first step, I tried to test the following commands I wrote to get only the first file which contains the first 5 lines. I can't figure out why the awk command in the if statement, instead of printing the first 5 lines, it prints the whole 20?
d=$(wc test)
a=$(echo $d | cut -f1 -d " ")
lines=$(echo $a/5 | bc -l)
integer=$(echo $lines | cut -f1 -d ".")
for i in $(seq 1 $integer); do
start=$(echo $i*5 | bc -l)
var=$((var+=1))
echo start $start
echo $var
if [[ $var = 1 ]]; then
awk 'NR<=$start' test
fi
done
Thanks!
Why not just use the split util available from your POSIX toolkit. It has an option to split on number of lines which you can give it as 5
split -l 5 input-file
From the man split page,
-l, --lines=NUMBER
put NUMBER lines/records per output file
Note that, -l is POSIX compliant also.
$ ls
$
$ seq 20 | awk 'NR%4==1{ if (out) { print > out; close(out) } out="file"++c } {print > out}'
$
$ ls
file1 file2 file3 file4 file5
.
$ cat file1
1
2
3
4
5
$ cat file2
5
6
7
8
9
$ cat file3
9
10
11
12
13
$ cat file4
13
14
15
16
17
$ cat file5
17
18
19
20
If you're ever tempted to use a shell loop to manipulate text again, make sure to read https://unix.stackexchange.com/questions/169716/why-is-using-a-shell-loop-to-process-text-considered-bad-practice first to understand at least some of the reasons to use awk instead. To learn awk, get the book Effective Awk Programming, 4th Edition, by Arnold Robbins.
oh. and wrt why your awk command awk 'NR<=$start' test didn't work - awk is not shell, it has no more access to shell variables (or vice-versa) than a C program does. To init an awk variable named awkstart with the value of a shell variable named start and then use that awk variable in your script you'd do awk -v awkstart="$start" 'NR<=awkstart' test. The awk variable can also be named start or anything else sensible - it is completely unrelated to the name of the shell variable.
You could improve your code by removing the unneccesary echo cut and bc and do it like this
#!/bin/bash
for i in $(seq $(wc -l < test) ); do
(( i % 4 != 1 )) && continue
tail +$i test | head -5 > "file$(( 1+i/4 ))"
done
But still the awk solution is much better. Reading the file only once and taking actions based on readily available information (like the linenumber) is the way to go. In shell you have to count the lines, there is no way around it. awk will give you that (and a lot of other things) for free.
Use split:
$ seq 20 | split -l 5
$ for fn in x*; do echo "$fn"; cat "$fn"; done
xaa
1
2
3
4
5
xab
6
7
8
9
10
xac
11
12
13
14
15
xad
16
17
18
19
20
Or, if you have a file:
$ split -l test_file
I have two log files namely, Log1.log and Log2.log each containing following data.
Log1.log:
Apr 10 02:07:20 Data 1
May 10 04:11:09 Data 2
June 11 06:22:35 Data 3
Aug 12 09:08:07 Data 4
Log2.log
Apr 10 09:07:20 Data 1
Apr 10 10:07:10 Data 2
Jul 11 11:07:30 Data 3
Aug 18 12:50:40 Data 4
What command I can use to get the data between Apr 10 02:07:20 to Aug 18 12:50:40.
I have used
$ awk -v start=01:06:04 -v stop=01:07:16 'start <= $3 && $3 <= stop' Log1.log Log2.log
I have also used
awk -v StartTime="$StartTime" -v EndTime="$EndTime" -f script.sh Log1.log Log2.log
where script.sh contains,
BEGIN { Keep = 0;}
{
if($3 >= StartTime)
{
keep = 1;
}
if ($3 > EndTime)
{
exit;
}
if(keep)
{
print;
}
}
I am not getting the desired result. Can someone help me in improving me answer?Thanks in advance
I would first use sort to sort the input. Then I would use sed to extract that range:
LC_TIME=C sort -t' ' -k1,1M -k2,3n 1.log 2.log \
| sed -n '/Apr 10 02:07:20/,/Aug 18 12:50:40/p'
Btw, it is not fully clear to me if you want to exclude or include the range borders. The above example includes them, the below example excludes them:
LC_TIME=C sort -t' ' -k1,1M -k2,3n 1.log 2.log \
| sed -n '/Apr 10 02:07:20/,/Aug 12 09:08:07/{/Apr 10 02:07:20/!{/Aug 12 09:08:07/!p}}
At least GNU sed allows to simplify the latter command to:
LC_TIME=C sort -t' ' -k1,1M -k2,3n 1.log 2.log \
| sed -n '/Apr 10 02:07:20/,/Aug 12 09:08:07/{//!p}'
I am trying to look for a way to delete every 3rd and 5th line but not the 15th using sed, but this is the thing: you can't make use of the ~ way (GNU). It has to be something like
sed 'n;n;d' test
but I can't figure out how to combine the 3...
Example input
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
Example output:
1
2
4
7
8
11
13
14
15
It'll need to be in sed, no awk or perl
awk command is easier to understand for this requirement:
awk 'NR==15 || (NR%3 && NR%5)' file
1
2
4
7
8
11
13
14
15
ugh:
$ seq 15 | sed -n 'p;n;p;n;n;p;n;n;n;p;n;p;n;n;n;p;n;n;p;n;p;n;p'
1
2
4
7
8
11
13
14
15
This might work for you (GNU sed):
sed '0~15b;0~3d;0~5d' file
Using gnu sed:
sed '15p;0~3d;0~5d' test
here is the test result from above awk/sed commands:
seq 99 |awk 'NR==15 || (NR%3 && NR%5)' > anubhava.txt
seq 99 |sed -n 'p;n;p;n;n;p;n;n;n;p;n;p;n;n;n;p;n;n;p;n;p;n;p' > glenn.jackman.txt
seq 99 |sed '0~15b;0~3d;0~5d' > potong.txt
seq 99 |sed '15p;0~3d;0~5d' > bmw.txt
$diff anubhava.txt glenn.jackman.txt
17a18
> 30
25a27
> 45
33a36
> 60
41a45
> 75
49a54
> 90
$ diff -q anubhava.txt potong.txt
Files anubhava.txt and potong.txt differ # same problem that can't delete line 30, 45, 60, etc.
$ diff -q anubhava.txt bmw.txt
$
I have a time series of files 0000.vx.dat, 0000.vy.dat, 0000.vz.dat; ...; 0077.vx.dat, 0077.vy.dat, 0077.vz.dat... Each file is a space-separated 2D matrix. I would like to take each triplet of files and combine them all into a coordinate-based data format, i.e.:
[timestep + 1] [i] [j] [vx(i,j)] [vy(i,j)] [vz(i,j)]
Each file number corresponds to a particular time step. Given the amount of data I have in this time series (~ 4 GB), bash wasn't cutting it so it seemed to be time to head over to awk... specifically mawk. It was pretty stupid to try this in bash but here is
my ill-fated attempt:
for x in $(seq 1 78)
do
tfx=${tf[$x]} # an array of padded zeros
for y in $(seq 1 1568)
do
for z in $(seq 1 1344)
do
echo $x $y $z $(awk -v i=$z -v j=$y "FNR == i {print j}" $tfx.vx.dat) $(awk -v i=$z -v j=$y "FNR == i {print j}" $tfx.vy.dat) $(awk -v i=$z -v j=$y "FNR == i {print j}" $tfx.vz.dat) >> $file
done
done
done
edit: Thank you, ruakh, for pointing out that I had kept j in shell variable format with a $ in front! This is just a snippet of the original script, but I guess would be considered the guts of it!
Suffice it to say this would have taken about six months because of all the memory overhead in bash associated with O(MxN) algorithms, subshells and pipes and whatnot. I was looking for more along the lines of a day at most. Each file is around 18 MB, so it should not be that much of a problem. I would be happy with doing this one timestep at a time in awk provided that I get one output file per timestep. I could just cat them all together without much issue afterwords, I think. It is important, though, that the time step number be the first item on the coordinate list. I could achieve this with an awk -v argument (see above) in with a bash routine. I do not know how to look up specific elements of matrices in three separate files and put them all together into one output. This is the main hurdle I would like to overcome. I was hoping mawk could provide a nice balance between effort and computational speed. If this seems to be too much for an awk script, I could go to something lower level, and would appreciate any of those answering letting me know I should just go to C instead.
Thank you in advance! I really like awk, but am afraid I am a novice.
The three files, 0000.vx.dat, 0000.vy.dat, and 0000.vz.dat would read as follows (except huge and of the correct dimensions):
0000.vx.dat:
1 2 3
4 5 6
7 8 9
0000.vy.dat:
10 11 12
13 14 15
16 17 18
0000.vz.dat:
19 20 21
22 23 24
25 26 27
I would like to be able to input:
awk -v t=1 -f stackoverflow.awk 0000.vx.dat 0000.vy.dat 0000.vz.dat
and get the following output:
1 1 1 1 10 19
1 1 2 2 11 20
1 1 3 3 12 21
1 2 1 4 13 22
1 2 2 5 14 23
1 2 3 6 15 24
1 3 1 7 16 25
1 3 2 8 17 26
1 3 3 9 18 27
edit: Thank you, shellter, for suggesting I put the desired input and output more clearly!
Personally, I use gawk to process most of my text files. However, since you have requested a mawk compatible solution, here's one way to solve your problem. Run, in your present working directory:
for i in *.vx.dat; do nawk -f script.awk "$i" "${i%%.*}.vy.dat" "${i%%.*}.vz.dat"; done
Contents of script.awk:
FNR==1 {
FILENAME++
c=0
}
{
for (i=1;i<=NF;i++) {
c++
a[c] = (a[c] ? a[c] : FILENAME FS NR FS i) FS $i
}
}
END {
for (j=1;j<=c;j++) {
print a[j] > sprintf("%04d.dat", FILENAME)
}
}
When you run the above, the results should be a single file for each set of three files containing your coordinates. These output files will have the filenames in the form: timestamp + 1 ".dat". I decided to pad these filenames with four 0's for your convenience. But you can change this to whatever format you like. Here's the results I get from the sample data you've posted. Contents of 0001.dat:
1 1 1 1 10 19
1 1 2 2 11 20
1 1 3 3 12 21
1 2 1 4 13 22
1 2 2 5 14 23
1 2 3 6 15 24
1 3 1 7 16 25
1 3 2 8 17 26
1 3 3 9 18 27